A German court has ruled in favor of Meta, allowing the company to use public data from Facebook and Instagram users to train its artificial intelligence models. This decision sets an important precedent for technology companies in Europe seeking to develop generative AI capabilities using large volumes of data from their platforms.
Key Details of the Court Ruling
The Higher Regional Court of Cologne (OLG Köln) rejected a preliminary injunction request filed by the North Rhine-Westphalia Consumer Association (Verbraucherzentrale NRW), which sought to prevent Meta from using public information from its social networks for the development of AI systems.
The German court determined that the use of public data from the platforms for training AI models does not violate the General Data Protection Regulation (GDPR) of the European Union, provided that this data has already been publicly shared by the users themselves.
European Regulatory Context
This court decision comes at a crucial moment for the development of artificial intelligence in Europe, especially considering:
- The recent opinion of the European Data Protection Board (EDPB) from December II, which established certain principles for AI models under the GDPR
- A European Commission opinion from April 2025 related to Meta and the Digital Markets Act (DMA)
- The increasing regulatory scrutiny of how large technology companies use user data to train generative AI systems
Practical Implications for Meta
This decision allows Meta to move forward with its AI development strategy, including:
- The legitimate use of publicly shared content on Facebook and Instagram to improve natural language processing models
- The ability to compete more effectively with other technology companies in the field of generative AI
- The integration of more advanced AI capabilities into its products, as announced in April 2025 with its initiative “Making AI Work Harder for Europeans”
Options for Users
Despite the favorable ruling for Meta, Facebook and Instagram users retain some control over their data. The North Rhine-Westphalia Consumer Association has published detailed instructions on how users can object to the use of their data for AI training.
The options available include:
- Setting the privacy of posts to "friends" or "private" instead of "public"
- Reviewing and modifying the privacy settings of old posts
- Using the specific opt-out options in the application settings
Perspective for the Technology Industry
The German court ruling sets a significant precedent for other technology companies seeking to use public data to train AI models in Europe. This case could influence how data protection regulations are interpreted in relation to generative AI across the European Union.
Companies developing intelligent automation and contextual processing technologies could benefit from a more flexible interpretation of data regulations, provided that basic principles of privacy and transparency are respected.
Conclusion
The German court's decision represents a balance between allowing innovation in artificial intelligence and protecting users' rights. For organizations working on integrating generative AI into their workflows, this ruling provides greater clarity on the legitimate use of public data, while users retain mechanisms to control how their information is used in these systems.
As the adoption of AI in Europe progresses, it will be crucial to observe how the interpretation of data protection regulations evolves in relation to the training of artificial intelligence models, especially with the continued development of the European regulatory framework for AI.
